Enterprise AI Infrastructure Becomes the New Battleground
Enterprise AI infrastructure is the collection of data platforms, pipelines, governance tools, and runtime environments that turn raw corporate data into features and signals that AI models can understand, trust, and use at scale to support business decisions and automation. As AI adoption accelerates, the fight has shifted away from models alone toward the data platform competition underneath. Databricks and ClickHouse are both racing to become the core layer between raw data and AI systems, defining how organizations ingest, prepare, analyze, and serve information for models and agents. Their rapid growth signals that enterprises are no longer satisfied with a traditional modern data warehouse focused on reporting. Instead, they want unified platforms that support AI data engineering, real-time analytics, and governed model serving on the same infrastructure, reducing tool sprawl and operational friction.
ClickHouse’s $250M ARR Signals Demand for Fast AI Analytics
ClickHouse’s recent milestone shows how critical fast analytics has become for AI workloads. The company has “surpassed USD 250 million (approx. RM1,150 million) in annualised revenue run rate, tripling year-over-year,” according to The AI Insider. Its columnar, open-source database is designed for high-speed analytical queries on massive datasets, a pattern that fits AI agents that must scan event logs, telemetry, or user interactions in real time. More than 4,000 customers, including Anthropic, Meta, and Capital One, are already standardizing on ClickHouse Cloud as their managed service. The company is also buying its way deeper into the AI stack, acquiring six startups such as Langfuse, which tracks and evaluates AI agent performance. With IPO plans signaled and a valuation of USD 15 billion (approx. RM69 billion), ClickHouse is positioning itself as the analytics nerve center that powers AI-first applications rather than a simple database.
Databricks’ $134B Valuation Anchors the AI Data Engineering Stack
Databricks has emerged as a flagship for enterprise AI infrastructure by turning its Lakehouse platform into the default choice for AI data engineering. The company closed more than USD 7 billion (approx. RM32 billion) in fresh funding, including USD 5 billion (approx. RM23 billion) in equity at a USD 134 billion (approx. RM614 billion) valuation, highlighting investor belief that this layer is where long-term value sits. Databricks blends Apache Spark, Delta Lake, MLflow, and Unity Catalog into a single environment where teams can build scalable ETL, govern data, and operationalize AI models. Gartner notes that 80% of enterprise data is spread across multiple platforms, and Databricks targets exactly this fragmentation with its unified Lakehouse architecture. By February, its annual revenue run rate had reached USD 5.4 billion (approx. RM25 billion), with fourth-quarter revenue growing more than 65% year over year, displaying unusual acceleration at its scale.
From Traditional Warehouses to Unified AI Data Platforms
The rise of Databricks and ClickHouse reflects a structural shift away from traditional data warehousing toward unified platforms tuned for AI workloads. Conventional warehouses were optimized for batch reporting, not continuous feature generation, streaming pipelines, or AI agent feedback loops. Businesses now need infrastructure that can handle both structured ERP-style data and unstructured sources such as documents and messages in one place. Databricks’ Lakehouse targets this need with a single platform for storage, compute, analytics, and ML lifecycle management. ClickHouse focuses on ultra-fast analytical queries on large-scale event and log data, a sweet spot for AI agents that depend on fresh context. Together, they show that the modern data warehouse is evolving into a broader enterprise AI infrastructure layer, where data engineering, governance, observability, and model serving converge rather than exist as separate tools and teams.
Different Paths, Same Goal: Owning the AI Data Layer
Although their starting points differ, Databricks and ClickHouse are aiming at the same prize: control of the data layer that makes AI useful in production. Databricks grew from big-data processing into a Lakehouse that unifies analytics and AI on one platform, supported by strong cloud partnerships and consulting ecosystems. ClickHouse began as a high-performance analytical database and is now expanding through acquisitions and managed cloud services toward an AI-aware analytics platform. For enterprises, the competition promises faster innovation but also raises strategic choices about lock-in, architecture, and skills. AI data engineering and modern data warehouse capabilities are converging into a single stack that prepares, governs, and serves data for models, automation, and agentic AI. The winner will likely be the platform that best simplifies this complexity while giving companies reliable control over their most valuable asset: their data.